Both bipolar mania and depression are characterized by mood instability and marked behavioral changes that appear to involve dysfunction of the anterior limbic network (ALN), a group of brain regions involved in emotional regulation and modulation. Functional MRI (fMRI) studies suggest that increased activity in the ventrolateral prefrontal cortex (VLPFC) and other regions modulates ALN structures involved in emotional expression to inhibit overt manifestations of mania and depression. Affective episodes represent a failure of these compensatory mechanisms, marked by decreased VLPFC activity and concomitant increases in activation of the amygdala and portions of the striatum. These neurofunctional changes are accompanied by a spectrum of neurochemcial changes including increased prefrontal glutamate and decreased concentrations of prefrontal N-acetyl-aspartate a marker of neuronal integrity. ALN dysregulation is also marked by abnormalities in neuronal metabolism and of the phosphatidylinositol cycle (PI-cycle), a second messenger cascade with multiple downstream neuronal effects. The therapeutic effects of lithium in both mania and depression have been linked to its actions on the PI-cycle;elevated concentrations of myo-inositol in affective patients are normalized with lithium treatment prior to amelioration of behavioral symptoms. As part of grant R01MH078043 we are obtaining fMRI and magnetic resonance spectroscopy (MRS) scans from manic patients at baseline, as well as one and eight weeks after beginning lithium. In this Revision Application, we propose to study a matched cohort of depressed bipolar patients. Because identical scanners and imaging methodologies will be employed, we will be able to compare neurochemical and neurofunctional effects of lithium treatment across mood state to address issues of state- vs. trait-related changes. Finally, we have added an exploratory specific aim to use a newer, systems biology approach to exploring the predictive value of baseline findings and early lithium effects. With these considerations in mind, we propose in this Revision Application to: 1) Use fMRI and MRS to measure neurofunctional and neurochemical differences between depressed and manic bipolar patients, and healthy controls at baseline, 2) Use fMRI and MRS to measure changes in neurofunctional and neurometabolic measures after one and eight weeks of lithium treatment, with the goal of refining neurophysiological models of lithium response across mood states, and 3) Identify MRS and fMRI markers and potential predictors of treatment response, using an exploratory Bayesian modeling methodology. PUBLIC HEALTH RELEVANCE: Although it is perhaps the most widely used bipolar medication in the world, the neurofunctional and neurochemical effects of lithium remain poorly understood. In this Revision Application we are proposing to add a depressed bipolar cohort to our ongoing fMRI and MRS study of manic patients to better understand the neurochemistry underlying the therapeutic effects of lithium across mood state. Ultimately, these findings could lead to more focused therapeutics and better predictors of lithium response.